Statins in pregnancyBMJ 2015; 350 doi: https://doi.org/10.1136/bmj.h1484 (Published 17 March 2015) Cite this as: BMJ 2015;350:h1484
- Françoise Haramburu, director,
- Amélie Daveluy, consultant,
- Ghada Miremont-Salamé, associate director
- Correspondence to: F Haramburu
The linked study (doi:10.1136/bmj.h1035) by Bateman and colleagues1 is an important contribution to our knowledge about the teratogenic potential of statins in early pregnancy. It is a well designed epidemiological study of pregnancy and childbirth among a large cohort of US women enrolled in the US Medicaid programme between 2000 and 2007.
The authors analysed 886 996 pregnancies that ended with a liveborn infant, representing approximately 40% of all births in the United States during the study period. They used sophisticated analytical techniques, including high density propensity scoring, to compare otherwise similar women who did or did not use a statin during their first trimester. Propensity scoring helps minimise confounding by adjusting for the many important differences between exposed and unexposed women, such as prevalence of diabetes (which is higher among women taking statins) and use of other drugs. These techniques are rarely used in studies of medicines in pregnant women and are an important strength of the new work.
Bateman and colleagues’ study evaluates the teratogenic potential of statins taken in the first trimester of pregnancy. The primary outcome was any major congenital malformation, but the authors also looked for organ specific malformations. While unadjusted results suggested a slight increase in the risk for any malformation associated with statin use (relative risk 1.79, 95% confidence interval 1.43 to 2.23), this increase disappeared in adjusted analyses that controlled for a wide range of potential confounders (1.07, 0.85 to 1.37).
The study was conducted in a cohort of women on low incomes, all of whom were eligible for government funded health insurance, Medicaid. Generalisability to other more affluent populations may be limited, but the demographic profile of the studied population arguably represents a “worst case scenario” for pregnancy outcomes. If results are reassuring among those on lowest incomes, they are likely to be equally reassuring for women living in a more favourable social and economic environment.
Despite using sophisticated methodology, database studies such as this one share several limitations, including the classification of statin use based on dispensed prescriptions and not consumption, and analyses confined to data available in the chosen database. The precise time of use of statins during pregnancy is difficult to assess from data on dispensed prescriptions. Misclassification can easily occur if, for example, adherence is poor (a common problem during long term primary prevention). As with most studies on drug use during pregnancy, Bateman and colleagues’ study does not have all the answers. Questions remain about statins and other outcomes related to congenital malformation (medical terminations, miscarriages, and intrauterine fetal deaths) and longer term effects on the neonate and beyond.
What are the practical implications of this new work for clinical practice, including the evaluation of risk among women taking statins who are pregnant or planning a pregnancy?
Statins are a preventive treatment for long term cardiovascular complications, seldom used in women of childbearing age (just 0.13% of women in this cohort were taking statins). Despite some pharmacokinetic differences, all statins share a common mechanism of action, so combining them in more than 1000 pregnancies makes pharmacological sense. The authors found no evidence of increased risk associated with statins, reinforcing similar results from most previous studies,2 3 4 5 6 7 8 except one.9 10
The US Food and Drug Administration currently designate statins in pregnancy as category X, or contraindicated. Statins in pregnancy are also contraindicated in Europe and in many other countries worldwide. Regulators will now have to update their information and potentially revisit a blanket ban. Women who unintentionally use statins in the early weeks of an unidentified or unplanned pregnancy can now be reassured about the risk. But how should treatment be managed once pregnancy is confirmed? In most women an interruption to treatment during pregnancy and breast feeding will not have serious consequences for the mother’s health, and despite Bateman and colleagues’ largely negative findings, interrupting treatment during pregnancy remains the best option. The benefits of statins are not established among women of childbearing age. Statin treated women who are planning a pregnancy can be advised to continue their treatment until pregnancy is confirmed.
Bateman and colleagues’ study does not consider the use of statins later in pregnancy so cannot inform choices beyond the first trimester. We have limited information about the safety of statins used “off label” for the prevention of pre-eclampsia, an indication currently under study.4 7 8 More information is also needed about safety outcomes beyond teratogenicity and the possibility of harm to the fetus or neonate from prolonged exposure to statins throughout pregnancy.
Bateman and colleagues’ well designed study is a welcome addition to the literature informing risk assessment during pregnancy. However, single studies are never enough. Teratogenic risks are notoriously difficult to detect, and absence of risk is even harder to establish with confidence. Women and their providers must make informed use of all available data when making decisions about treatment during pregnancy. Tragic outcomes from drug use during pregnancy such as those due to thalidomide or diethylstilbestrol must not be forgotten. It would be unwise to make wholesale changes to recommendations and advocate wider use of statins during pregnancy on the basis of reassuring results from one or more single studies, however well conducted. The latest study along with its predecessors cannot eliminate uncertainty, a now familiar aspect of all therapeutic discussions about the risks of drug use in pregnancy.
Cite this as: BMJ 2015;350:h1484
Competing interests: We have read and understood the BMJ policy on declaration of interests and declare the following interests: none.
Provenance and peer review: Commissioned; not externally peer reviewed.